For the past three decades, there have been numerous applications for single, double, and triple resonance circuits in NMR of solid samples where high rf magnetic field strength B.sub.1 at high field B.sub.0 is required with minimal signal loss for high sensitivity. In U.S. Pat. No. 4,968,938, Pines and Samoson disclose a technique, Dynamic Angle Spinning (DAS), for improving resolution in quadrupolar solids. The DAS technique requires rapid reorientation of the axis of a rapidly spinning sample. Variable Angle Spinning (VAS) probes have been commercially available since 1983 for related techniques for quadrupolar nuclides as described by Oldfield. Other techniques requiring rapid reorientation of the spinning axis (Switched Angle Spinning, SAS) were described in 1984 by Terao and others. More recently, VAS and SAS techniques have been shown to be quite useful in biological membranes, especially where both the wideline spectra with the tissue oriented at 90.degree. with respect to B.sub.0 and the High Resolution Magic Angle Spinning (HR-MAS) spectra are needed to determine the structure. However, because of the technical difficulties involved in producing a multinuclear SAS or DAS probe capable of high B.sub.1 and B.sub.2 (and possibly B.sub.3) with high sensitivity, only a handful of such probes have been produced, even though numerous applications have been identified for at least fifteen years.
For experiments on solid samples at high B.sub.0, typical voltages across the solenoidal sample coil are 2 to 6 kV. Broadband (multinuclear) triple-resonance circuits as disclosed by Doty in U.S. Pat. No. 5,424,645 have provided multinuclear tuning capability on a Low-Frequency (LF) and Mid-Frequency (MF) channel with high-power .sup.1 H decoupling for various NMR experiments such as REDOR and double CP. Normally, a sample spinner similar to the one described by Doty et al in U.S. Pat. No. 5,508,615 (note the extensive list of typographical corrections) is utilized, as it has the symmetry needed to permit rapid reorientation. Some of the probe requirements are reviewed briefly by Doty in `Solid State NMR Probe Design`, The NMR Encyclopedia, Vol. 6, Wiley Press, 1996. Several copending applications describe methods of improving resolution and decoupling efficiency in HR-MAS.
For the past five decades, NMR probes for solid samples have almost exclusively utilized solenoidal sample coils, although fixed-tuned transverse coils for the high frequency (HF) in combination with solenoids for the LF are shown in a copending application to offer advantages in many CPMAS applications. In this case, the HF coil may be approximately tuned by fixed capacitors within a few millimeters of the coil to the desired HF, and the voltage standing wave ratio (VSWR, defined in terms of the reflection coefficient in the usual manner) on the balanced transmission line driving the HF coil may be as low as 10 to 40 for a characteristic HF line impedance of about 120 .OMEGA. and still achieve adequate tuning range. Peak differential HF voltages are normally under 1.4 kV for this low-inductance coil arrangement. The LF/MF solenoid, on the other hand, cannot be broad-banded and/or double tuned without VSWR's typically in the range of 100 to 500 for a balanced LF transmission line impedance of about 50 .OMEGA.. Peak differential LF/MF coil voltages are usually in the range of 2.2 to 5 kV, and pulsed rf currents are typically around 50 amps.
Clearly, one of the most difficult technical problems has been the flexible, high-power rf leads, where conventional tranmission-line analysis is of limited utility because of the high VSWR. A variety of approaches have been tried, but all have been unsatisfactory from a practical and performance perspective for multinuclear applications. Sliding contacts from a variety of materials, arranged in an arc either beside or below the spinner assembly, may allow reduced variation in lead capacitance as the angle is changed, but the mechanism's capacitance and contact resistance make this approach quite unsatisfactory, as one might well imagine with UHF VA products well over 10.sup.5 VA. The least satisfactory prior art from an rf performance perspective used `watch-spring` leads to permit millions of flips without fatiguing. Most prior art utilized several parallel leads of flattened silver-plated fine copper braid without polymeric insulation. The leads then must be well spaced to prevent arcing as their relative positions change throughout sample reorientation. The primary problem with this approach is that the nominal characteristic impedance of the leads must be high (typically over 220 .OMEGA.) to insure adequate separation, and the braid usually fatigues after fewer than a thousand flips. Moreover, the attenuation constant is generally greater than 0.15 dB/m at 200 MHz, and the high characteristic impedance results in increased voltage transformation, VSWR, and losses.
Lightly twisted stranded wires of several hundred strands of #40 AWG silver-plated copper wire will perform satisfactorily for perhaps a thousand flips near room temperature if insulated with polypropylene or teflon foam about 0.5 to 1.5 mm thick, as the wires may be bundled together using a teflon string wrap to produce a balanced transmission line of sufficiently low characteristic impedance (under 120 .OMEGA.), high voltage (HV) rating above 2.2 kV, high propagation velocity factor (preferably greater than 85%), and low attenuation constant (under 0.1 dB/m at 200 MHz). The ultra low attenuation constant is needed even with short leads because the signal loss is multiplied by the VSWR. The primary disadvantage of this approach is the limited lifetime, owing to fatiguing of strands within the bundles. Moreover, because of the thickness of the insulation, it is too brittle below about -50.degree. C., and operation is frequently needed from -140.degree. C. to +200.degree. C.
Another prior art technique utilized two separate, parallel strips of thin polyimide film (such as Kapton or Regulus) about 0.05 mm thick with copper cladding about 0.06 mm thick on one side. The strips, each 5 to 10 mm wide, were attached separately to the rf sample coil terminations and spaced 10 to 20 mm apart to prevent shorting as they flexed during changes in the spinner angle, as shown in FIG. 1. Annealed, rolled copper cladding is generally preferred over electro-deposited copper for low-loss rf conductors, as microscopic defects in electro-deposited conductors may drastically reduce electron mean free paths and hence conductivity. After a few flex cycles, the annealed copper becomes work hardened and its yield strength increases substantially without significantly affecting conductivity. However, work-hardened copper still has only moderate yield strength, and stresses must be limited to about half of the yield stress to permit millions of flex cycles without failure. Hence, for 0.06 mm copper foil, the minimum flex radius for very long life is about 20 mm--or perhaps a flex radius cycle between 20 and 10 mm. Since the attenuation constant increases rapidly with decreasing frequency below the point at which the foil thickness is less than four times the rf skin depth (and the characteristic impedance also increases, but more slowly), the copper thickness usually must be greater than 0.04 mm, as low-loss performance below 50 MHz is often needed. Moreover, greater conductor thickness is desired even at UHF to handle the heating from 10 ms 50 A rf pulses; and increasing the thickness from 4 to 10 skin depths reduces attenuation constant by about 30% over a rather wide range of twinline impedances.
The laminate of this prior art is similar to that used in multi-conductor controlled-impedance ribbon cables commonly found in the computer and communications industries. Many electronic devices employ flexible multi-conductor cables consisting of parallel arrays of conductor strips on one or both sides of a polyimide substrate for power, logic, audio, and rf applications at low and moderate voltages. However, the low-voltage, high-loss fully shielded moderately flexible 50 .OMEGA. rf strip lines commonly used in communications, such as shown by Borokowski et al in U.S. Pat. No. 5,154,625, are completely unsatisfactory for the DAS NMR probe coil leads. The foil shielding makes this `flexible` line rather rigid from the perspective of the DAS probe requirements, limits the propagation factor to a maximum of about 55%, and limits attenuation constant to 2 to 20 dB/m at 200 MHz, depending on substrate thickness. Moreover, the high-impedance separate strips of FIG. 1 are essentially unsatisfactory at high B.sub.0, where low-inductance multinuclear rf coils must be used. The primary disadvantages of the prior art shown in FIG. 1 are its high and variable characteristic impedance (typically over 250 .OMEGA.) and moderate attenuation constant. While the ultra-low-loss flexible HV rf transmission lines disclosed herein, generically denoted as flexlines, should have applications outside of NMR probes (especially if scaled down for a lower voltage rating), the absence of an external rf shield will limit their utility to applications where external shielding is not needed or is already provided by other means.